University of Joensuu, PhD Dissertations in Biology No: 48
Effects of mining industry waste waters on fish in the Kostomuksha area,
NW Karelia, Russia
Victoria Tkatcheva
ACADEMIC DISSERTATION
To be presented, with the permission of the Faculty of Biosciences of the University of Joensuu, for public criticism in the Auditorium B1 of the University, Yliopistokatu 7, on 31st August, 2007, at 12 noon
Pre-examiners Professor Aimo Oikari
Department of Biological and Environmental Science University of Jyväskylä, Finland
Docent Pekka Vuorinen
Finnish Game and Fisheries Research Institute Helsinki, Finland
Examiner
Professor Karl Erik Zachariassen Department of Biology
The Norwegian University of Science and Technology Trondheim, Norway
Supervisors
Professor Ismo J. Holopainen, Professor emeritus Heikki Hyvärinen
and Academy Professor Jussi V. K. Kukkonen
Faculty of Biosciences, University of Joensuu, Finland
University of Joensuu Joensuu 2007
Julkaisija Joensuun yliopisto, Biotieteiden tiedekunta PL 111, 80101 Joensuu
Publisher University of Joensuu, Faculty of Biosciences P.O.Box 111, FI-80101 Joensuu, Finland Toimittaja FT Heikki Simola
Editor Dr
Jakelu Joensuun yliopiston kirjasto / Julkaisujen myynti PL 107, 80101 Joensuu
puh. 013-251 2652, fax 013-251 2691 email: joepub@joensuu.fi
Distribution Joensuu University Library / Sales of publications P.O.Box 107, FI-80101 Joensuu, Finland
tel. +358-13-251 2652, fax +358-13-251 2691 email: joepub@joensuu.fi
Verkkojulkaisu http://joypub.joensuu.fi/joypub/faculties.php?selF=11 väitöskirjan yhteenveto-osa; toim. Markku A. Huttunen and Tomi Rosti
ISBN 978-952-458-981-9 (PDF)
Internet versionhttp://joypub.joensuu.fi/joypub/faculties.php?selF=11 summary of the dissertation; ed. by Markku A. Huttunen and Tomi Rosti
ISBN 978-952-458-981-9 (PDF)
Sarjan edeltäjä Joensuun yliopiston Luonnontieteellisiä julkaisuja (vuoteen 1999) Predecessor Univ. Joensuu, Publications in Sciences (discontinued 1999)
ISSN 1795-7257 (printed); ISSN 1457-2486 (PDF) ISBN 978-952-458-980-2 (printed)
Joensuun Yliopistopaino 2007
Victoria Tkatcheva
Effects of mining industry waste waters on fish in the Kostomuksha area, NW Karelia, Russia. – University of Joensuu, 2007, n pp.
University of Joensuu, PhD Dissertations in Biology, No. 48.
ISSN 1795-7257 (printed), ISSN 1457-2486 (PDF)
ISBN 978-952-458-980-2 (printed), ISBN 978-952-458-981-9 (PDF)
Key words: environment, perch, roach, rainbow trout, heavy metal concentration, liver, gills, histological structure, enzyme activities, lipid composition.
The problem of detecting the effects of heavy metals on nature is receiving considerable attention in North America, Europe and Russia. This thesis studies the morphological, physiological and ecotoxicological effects of elevated metal concentrations and changed water quality on individual fish and their populations.
The changes are attributable to mining industry wastewaters in the forest lakes of the Kenti river system (North-west Karelia, Russia). Two fish species were studied, the piscivorous perch (Perca fluviatilis) and the detritivorous- herbivorous roach (Rutilus rutilus).
Concentrations of copper, zinc, cadmium, mercury and chromium were analyzed in the water, sediment and fish from the Kenti river system. The combined effects of the alkali metals, lithium and potassium, and of lithium alone, were investigated in laboratory experiments using juvenile rainbow trout (Oncorhynchus mykiss). Fish liver and gill structure, lipids and enzyme activities were studied in response to metal concentrations, seasonal factors, and the environment.
The release of wastewater caused changes in the water quality and sediment composition. In addition to the abiotic factors in their habitat, the feeding habits and life style of the fish were important factors in their metal uptake. A lower mercury concentration in fish was registered in spring before snowmelt. In spite of the low metal bioavailability caused by the high water hardness and pH in L. Poppalijärvi, chronic Cu exposure was pronounced due to its high concentration in water, sediments and roach. Physiological changes in the gill chloride cell structure imply an increase in osmoregulatory work in L. Poppalijärvi.
Cholesterol was shown to be the most adaptive compound in the gill membrane, both in the field and in the laboratory. Lithium may affect fish through diffusive sodium ions losses at the gills and by reduced enzymatic activity of the gills, possibly related to observed lower concentrations of free fatty acids and cholesterol in gill tissue. However, no destructive effects of Li on Na+,K+-ATPase activity were found in the presence of potassium.
4
Content
List of original publications... 5
1. Introduction... 6
2. Background... 6
2.1. Kenti river system and mining industry pollution... 6
2.2. Accumulation and toxicity of metals in fish ... 9
2.3. Effects of temperature, pH level, water hardness, and iron deposition on metal uptake and toxicity ... 12
2.4. Mechanism of metal toxicity in gills of freshwater organisms ... 13
3. Material and methods... 15
3.1 Field sampling ... 15
3.2. Laboratory experiments... 15
3.3. Analyses ... 16
3.3.1. Heavy metals ... 16
3.3.2. Histology and histochemical analyses... 16
3.3.3. Enzyme analyses ... 17
3.3.4. Lipid analysis ... 17
3.3.5. Osmolality and ionic concentrations of plasma ... 17
3.3.6. Apolipoprotein AI analysis ... 17
3.3.7. Statistics ... 17
4. Results... 18
4.1. Fish communities ... 18
4.2. Metals in the water and sediment... 18
4.3. Metal concentrations in the fish ... 20
4.4. Fish tissue damages... 22
4.4. Lipids in the gills... 22
4.5. Enzyme activity in gills... 25
4.6. Plasma ion concentrations under Li exposure... 25
4.7. Apolipoprotein AI under Li exposure ... 25
5. Discussion... 25
5.1. Metal concentration in the fish as affected by season, fish feeding behavior, water and sediment parameters ... 25
5.2. Liver damage, histology aspect... 29
5.3. Gill responses ... 29
5.4. Lipid responses in the gills... 30
6. Conclusions... 31
Future research needs... 32
Acknowledgements... 32
References... 33
5 List of original publications
This thesis is based on the following articles, which are referred to in the text by their Roman numerals:
I Tkatcheva V., Holopainen I.J., Hyvärinen H.. 2002. Effects of mining waste waters on fish in lakes of NW Russia. Verh. int. Ver. Limnol. 28, 484-487.
II Tkatcheva, V.G., Holopainen, I.J., Hyvärinen, H. 2000. Heavy metals in perch (Perca fluviatilis) from the Kostomuksha region (north-west Karelia, Russia). Boreal Environment Research 5, 209-220.
III Tkatcheva V., Holopainen, I J., Hyvärinen, H., Ryzhkov, L.P., Kukkonen J. 2004.
Toxic effects of mining effluents on fish gills in a subarctic lake system in NW Russia, Ecotoxicology and Environmental Safety, 57, 278-289.
IV Tkatcheva, V., Holopainen, I.J., Hyvärinen, H., Kukkonen, J.V.K. The response of rainbow trout gills to high potassium and lithium in water. Ecotoxicology and Environmental Safety (in press).
V Tkatcheva, V., Franklin, N.M., McClelland, G.B., Smith, R.W., Holopainen, I.J., Wood, C.M. Physiological and biochemical effects of lithium in rainbow trout. Manuscript accepted for publication (December 2006) in Archives of Environmental Contamination and Toxicology.
Publications are reprinted with permission from the publishers. The copyright for publications I is held by E. Schweizerbart'sche Verlagsbuchhandlung http://www.schweizerbart.de; II by Boreal Environment Research, Helsinki; III and IV by Elsevier; V by Springer.
Some unpublished results are also presented and discussed.
6 1. Introduction
Metals, such as mercury, cadmium, copper and zinc, form major types of toxic compounds that are released into many watercourses by the mining industry (Mason, 2002). Both air emissions and wastewaters are sources of metal pollution in the Kenti river system (Morozov, 1998;
Lozovik et al., 2001, Kalinkina et al., 2003). In the late 1990s, metal pollution of the Kenti river system caused by the mining industry was not investigated to any great extent, and the heavy metal contamination of its fish had not been studied. Nevertheless, one of the few studies from the area has shown that molluscs (Lymnaea stagnalis L. and Sphaerium sp.) had elevated levels of metallothionein in the upper part of the Kenti river system (Regerand, 1995), thus indicating exposure to potentially toxic concentrations of metals (Amiard et al., 2006). In addition, preliminary results from the years 1997-1998 showed that fish from the upper part of the water system had higher tissue concentrations of metals. The studies in this thesis were undertaken to investigate the environmental impacts of metal pollution caused by the release of wastewater into the Kenti river system by Kostomuksha Mining Plant (KMP).
The main purpose of this thesis was to find out how metals accumulate and affect feral fish in the Kenti river system.
In 2000, the project was initiated as an investigation of metal impact on the biota of the natural Kenti river system.
Holopainen et al. (2003a, b) explained changes in the biotic community from bacteria to fish. Perch and roach, which have different feeding habits and lifestyles, were chosen to demonstrate the differences in metal uptake and the interrelations with abiotic factors in the Kenti river system.
Concentrations of metal were studied in fish muscle and liver tissue (I, II) and in the gills (III). Since high concentrations of
metals do not imply that the metals have a toxic effect (Pettersen et al., 2002), toxicity of metals was mostly associated with vital physiological functions, such as enzyme activity, modifications in membrane lipid composition, and changes in tissue structures. The last part of this work focused on alkali metals, which were abnormally high in the Kenti river system.
The effects of alkali metals on the gills of juvenile rainbow trout (IV, V) were studied in the laboratory.
The objectives of this thesis were to find out:
1. The effects of metal pollution on individual fish and fish populations in successive lakes downstream from the Kostomuksha Mining Plant (I, II, III) 2. Possible changes in the structure of
fish liver and gills in response to metal concentrations (II, III).
3. Possible interaction between gill enzyme activity and lipid composition under wastewater impact, and the role of the lipids in seasonal and environmental adaptation (III).
4. The combined effects of the alkali metals, lithium and potassium, on gill structure, lipid composition and enzyme activity in juvenile rainbow trout in the laboratory (IV).
5. The effects of lithium on the physiology of juvenile rainbow trout in the laboratory (V).
2. Background
2.1. Kenti river system and mining industry pollution
The Kenti river system is an example of the beautiful, but vulnerable natural northern river systems of the Republic of Karelia, Russia. Historically, the study area belongs to Belomorskaja (White Sea) Karelia, where Elias Lönnrot collected the old Karelian poems, charms and conjurations that formed the basis for the
7 Finnish national epos, the Kalevala (The Kalevala, 2004).
One of the ecological problems in this area is the wastewater discharge from the Kostomuksha Mining Plant (KMP).
The production operations have severe effects on the environment in the form of air pollution (SO2 and dust) and wastewater emissions (Anonymous, 1995;
Kuharev et al., 1995; Fedorets et al., 1998;
Virtanen et al., 2000; Lozovik et al., 2001). The mining effluents have been deposited in the closed basin since 1982.
Since 1994, the overflow of wastewaters into the Kenti river system has been allowed. Formerly the depository was a natural lake, L. Kostamus (or Kostomuksha). Water from Lake Kostamus flows out to the northeast through the Kenti river system into the larger Kuito Lakes and the White Sea (see map in articles I-III). The contamination of this area has been studied since KMP started to use the upper lake of this system as a waste basin. Water quality is monitored by the Karelian Science Center (Northern Water Research Institute, Petrozavodsk, Russia) and by the Kainuu Regional Environment Centre (Finland).
The Kenti watercourse is a typical boreal lake-river system with low mineral content, soft brown water with a slightly acidic pH, and with a low average annual temperature (Lozovik et al., 2001;
Holopainen et al., 2003b). The main ion composition of the water is altered in the upper part of the Kenti system by a considerable rise in total nitrogen and in Mg, Ca, Na, Cl, SO4, K and Li, which has changed the water pH from acidic to alkaline and increased the hardness of the water. It seems that the effect of water quality on solubility and speciation will decrease the availability of metals for uptake by biota and thus help to make the metal pollution less severe (Lozovik et al., 2001; Platonov and Lozovik, 2003). It has been suggested that the wastewaters cause a threat to biodiversity, impair local fishing
and contaminate drinking water (e.g.
Kalinkina et al., 2003).
The effects of wastewater pollution on biota as described by Holopainen et al (2003b) are the following: firstly, the primary producers (including picocyanobacteria) and the fish in the lakes downstream from the mine had double the biomass compared to fish in the reference lake; secondly, zooplankton and zoobenthos biomasses are highest in L.
Kento, the lake in the middle of the outflow series; thirdly, the fish fauna is the same in all the lakes and is dominated by perch and roach. Both grow well, and the roach show the best growth in L.
Poppalijärvi, next to the waste basin. Since the phosphorus levels are equally low in all the studies of the lakes, the high nitrogen content (source: quarry explosives) in the Kenti river system might be partly responsible for the higher phytoplankton and picocyanobacteria biomasses, an important food for Ciliata, zooplankton and zoobenthos, which in turn are important food sources for the fish. The high biomass of benthic animals was assumed to be more important for the predominant roach than that of the pelagic zooplankton (Holopainen et al., 2007). The dominant fish, perch and roach, were considered to be suitable organisms for demonstrating metal accumulation.
The natural source of metals for the Kenti river system is the chemical weathering of rocks and soils. This territory is part of the Baltic or Scandinavian Shield area, and the ancient Precambrian silicate rocks are rich in metals, especially mercury (Startchev, 1985). Another source of metals in the lake water are the wastewaters from mining processes (Lozovik et al., 2001). For example, the following concentrations of metals have been measured from the precipitation basin in the liquid and solid phases (Lozovik et al., 2001): iron 270 mg l-1 in the liquid phase and 1500 mg l-1 in the solid phase, chromium 0.09 to 2.9 and zinc 0.08 to 1.2 mg l-1, in the liquid and
8 solid phases respectively. The alkali metals potassium and lithium are found in concentrations of 143 to 330, and 0.16 to 0.53 mg l-1, the divalent cations calcium in concentrations of 18 to 630 and magnesium in concentrations of 9 to 130 mg l-1, respectively. Two canals bring metals into the Kenti river system. The South canal flows directly into L.
Poppalijärvi and collects waters from the drainage area of the depository, including the flow from four rivers contaminated heavily by metals from the tailings of the open pit mine (Lozovik et al., 2001).
In addition to our own results from the field study, data from Morozov (1998), Virtanen and Markkanen (2000) and Lozovik et al. (2001) have been used, and they are presented in the chapter entitled
“Results” (Table 4). The water samples collected at the same time as the fish tissue samples were collected in the years 2000 and 2001.
The high concentrations of metals in the sediment of L. Poppalijärvi (Table 5) reflected the binding of metals by absorption and coprecipitation with hydrous Fe and Mn oxides, and with Ca and Mg carbonates (III). Although, immobilisation of metals seems to be very high in the lake next to the waste basin, mobilisation of metals is registered downstream in L. Kento. This is attributable to changes in the water conditions towards lower pH and low mineral content, including low complexation or chelation by DOC.
Perch (Perca fluviatilis L.) and roach (Rutilus rutilus L.) are common freshwater species found widely throughout Europe (Lappalainen et al., 2001). Both are food generalists and they have no special requirements for spawning areas. The investigation by Lappalainen et al. (2001) demonstrated that in the Tvärminne area in SW Finland, for over twenty years there was an increase in roach stocks (main food molluscs) in comparison to perch (main food macro-crustaceans and fish).
Eutrophication was suggested as the main
cause of the increase in the roach population. This agrees with the study of Persson (1983), according to which roach are adapted to utilising molluscs, and they can crush their food mechanically in the dark, whereas perch are visual hunters dependent on vision, as reported by Diehl (1988). Perch, being carnivorous predators of relatively large and mobile prey, expend a lot of energy per capture, whereas roach as an omnivorous species have a feeding strategy that minimises energy expenditure (Persson, 1983; Lappalainen et al., 2001).
In the Kenti river system, roach (comprising 35–51% of biomass) and perch (32–59%) were the most numerous fish (Holopainen et al., 2003b). The mean size of roach in L. Upper Kuito (the reference lake) appeared to be smaller than that in L. Kento (7 km from the depository) or in L. Poppalijärvi (3 km, from the mining plant). Yet young fish <10 cm were found in equal numbers in all lakes, suggesting normal reproduction. The growth rate of perch appeared to be similar in all three lakes in both 2000 and 2001.
On average, perch reached a length of 17 cm at the age of 4+ years. Roach grew well in both the impacted lakes, reaching a length of 18 cm at the age of 4+, but showed a lower growth rate in the reference lake, with a length of only 15 cm at the age of 4+ years. A preliminary stomach analysis of a small number of young perch from all the lakes suggested a diet of insect larvae (e.g., Ephemera vulgata) and pupae. A large number of Chydoriidae (Cladocera) was found in L.
Kento and L. Upper Kuito. The roach diet, however, appeared to differ between the lakes. Cladocera and Chironomidae predominated in L. Upper Kuito, but the larger Ephemera and Sphaeriidae (Pisidium sp.) were predominant in the two other lakes. This finding is in agreement with littoral zoobenthos samples, where the proportion of molluscs was highest in L.
Poppalijärvi (Holopainen et al., 2003a, 2007; Aroviita et al., 2006).
9 2.2. Accumulation and toxicity of metals in fish
Mercury (Hg) and cadmium (Cd) have no known biological function (Di Giulio et al., 1995), but are able to bioaccumulate in the food chain (Mason et al. 1994; 1999;
2000). They can be toxic in small quantities and are present at high levels even in the Arctic region, far from most anthropogenic sources (AMAP, 2002).
Mercury:
Mercury accumulation in fish depends on the following: its chemical form (the most toxic being methylmercury), the water temperature, chlorine ions, dissolved organic matter (DOM), dissolved organic carbon (DOC) and hardness (Grieb et al., 1990; Newman and Jagoe, 1994; Hamilton, 1995; Downs et al., 1998; Mason et al., 2000; Wood, 2001; Duffy and Zhang, 2002). Selenium (Se) can reduce the toxicity of Hg in aquatic organisms by competing for protein binding (Kai et al., 1995; Dorea et al., 1998; Nuutinen and Kukkonen, 1998). Methylmercury (MeHg) is mostly bound in protein via association with the thiol group in the muscle of the fish, while inorganic Hg is usually concentrated in the detoxifying organs, such as the kidney and the liver (Mason et al., 2000). The gills are not the dominant uptake routes for MeHg, and blood rapidly transfers and distributes MeHg to internal organs (Mason et al., 2000). Inorganic Hg accelerates the uptake of methylmercury by an unknown mechanism in the gills (Rodgers and Beamish, 1983; Wood, 2001). The loss mechanism for metalloids is depuration. Inorganic Hg is more effectively eliminated than methylmercury in clams (Inza et al., 1998).
For both forms of mercury, a failure in osmoregulation appears to be the proximate cause of acute toxicity (Heath, 1995). In freshwater fish the concentration- dependent decrease may occur in plasma Na+, Cl- concentrations and osmolality (Lock et al., 1981). These negative effects
have two separate mechanisms: an increase in plasma-membrane osmotic water permeability (at a level well below 96-h LC50), and an inhibition of gills Na+,K+- ATPase, which reduces Na+ uptake at concentrations close to the 96-h LC50 (in the range of 30-700 µg Hginorg l-1) (Lock et al., 1981; Spry and Wiener, 1991).
Mercury has a strong affinity for sulfhydryl groups, which explains its ability to inhibit branchial Na+,K+-ATPase (Mason et al., 2000).
The specific morphologic response of freshwater trout exposed to methylmercury was an increase in the frequency of mitotic figures in pavement cells, while seawater windowpane flounder that were chronically exposed to inorganic mercury showed chloride cell proliferation (Wood, 2001). Other non-specific responses could be swelling and hyperplasia of pavement cells, necrosis, excess secretion of mucus and mucous cell proliferation, epithelial rupture and fusion of lamellae. Fish exposed to high mercury levels may suffer damage to their gills, blindness and they may develop a reduced ability to absorb nutrients through the intestine. Mercury levels in fish are related to their age and size (Bodaly et al., 1993).
Cadmium:
Cadmium is taken up mainly through the intestine in fish and to a much lesser extentthrough their gills (AMAP, 2002).
Fish kidneys and gills are organs for depuration, and they are also areas where metals accumulate, either directly from the water or from the food (Mason et al., 2000). Dietary cadmium is initially bound to the albumin and the blood cells, and then to metallothionein in the liver and kidneys.
Cadmium is considered to be a respiratory toxicant (Hollis et al., 1999) and a nephrotoxicant for fish (Sangalang and Freeman, 1979). Lethal concentrations of cadmium in water may vary greatly, since Cd accumulation in fish depends on
10 water quality (pH, DOC etc., Table 1) and the sensitivity of the fish species (Hollis et al., 1999, 2001). Toxic Cd concentration is below 100 µg l-1 in freshwater fish (Wood, 2001). In fish gills, both Cd uptake and toxicity occur via the chloride cells (Wicklund-Glynn et al., 1994, Wood, 2001). Cadmium is an antagonist of Ca2+
uptake in fish gills, because the intracellular free Cd2+ ions competitively inhibit the basolateral Ca2+-ATPase (Reid and McDonald, 1988; Verbost et al., 1989;
Hollis et al., 2000). Uptake and plasma concentrations of Na+ and Cl- are less important, but may occur at higher threshold levels of Cd. These contribute to the damage to or death of chloride cells by both apoptosis and necrosis, and an associated loss of Na+,K+-ATPase activity.
The recent investigation by Niyogi and Wood (2004) has shown that branchial influxes of Ca2+ and Cd2+ occur through common pathways. In the analyses of inhibitor constants for branchial Cd2+
uptake by waterborne Ca2+, they indicated that the inhibition was three times more potent in yellow perch than in rainbow trout.
A respiratory effect during sublethal exposure was reported as a common morphological response, namely increased blood-to-water diffusion distance in the gills resulting from cellular proliferation, and hyperplasia of the chloride cells (Wendelaar Bonga and Lock, 1992, Hollis et al., 1999). In contrast to Hg, there is no metallothionein induction in the gills by Cd (Hollis et al., 2001).
Chromium:
Chromium is present in the water as various anions of Cr(III) and Cr(VI). The gills, liver, kidney and the digestive tract are the organs that accumulate the Cr at alkaline pH and in high environmental concentrations. At neutral pH only the gills accumulate Cr (Eisler, 1986). Adverse effects of chromium have been documented at 10.0 µg l-1 of Cr(VI) and 30
µg l-1 of Cr(III), in freshwater, for sensitive species (Eisler, 1986). No biomagnification of Cr has been observed in food chains, and its concentration is usually highest at the lowest trophic level.
Arillo et al. (1982) reported that only male rainbow trout showed changes in liver enzyme activities (0.2 ppm Cr(VI) for 6 months). The effects were more intensive when Ni and Cd ions were present. In trout, Cr(VI) uptake increased when 10 ppb of ionic Cd was also present in the solution (Calamari et al., 1982). In fish gills Cr(III) is directly coupled with the transfer of oxygen, and this reaction is more rapid at pH 6.5 than at alkaline pH (Van der Putte and Part, 1982; Eisler, 1986).
Copper and Zinc:
Essential metals such as copper (Cu) and zinc (Zn) are regulated by specific transport mechanisms in the gills and gut (Wood, 2001; Bury et al., 2003).
Copper is used in numerous enzymatic processes because of its capability to accept or donate ions (redox reactions) (Hochachka and Somero, 2002, Bury et al., 2003). However, Cu2+ is an ionoregulatory toxicant in fresh water, and pathophysiological effects are registered at total concentrations below 100 µg l-1 (Wood, 2001). Copper induces ionoregulatory dysfunction in the gills by inhibiting Na+ and Cl- influx (insensitive to [H+]) and the effects are similar to those of the more toxic silver and aluminum (Wood, 2001). Copper has a strong affinity for the sulfhydryl group, which explains its ability to inhibit branchial Na+,K+-ATPase activity both in vitro and in vivo (Pelgrom et al., 1995; Li et al., 1996). Copper may increase gill transcellular permeability, since K+ losses were disproportionately high relative to Na+ and Cl- losses (Lauren and McDonald, 1985). In addition, copper may induce lamellar damage and mucification in the gills (Wilson and Taylor, 1993).
11 According to Wood (2001), the blood pH level of fish was increased as a result of metabolic alkalosis in copper sublethal exposure, which may reflect an internal build-up of ammonia in the plasma and tissue, a phenomenon which has been implicated in reduced swimming performance (Lauren and McDonald, 1985; Taylor et al., 1996; Wang et al., 1998; Beaumont et al., 1995a,b).
Furthermore, the decreased aerobic swimming performance of the fish may reflect changes in gill morphology, such as epithelial lifting, lamellar fusion, necrosis and apoptosis caused by cortisol (Bury et al., 1998). This is in contrast to the necrosis that is caused by copper directly (Li et al., 1998; Bury et al., 1998, Wood, 2001). Repair during chronic exposure involved proliferation of chloride cells and mucous cells with an accompanying increase in the total Na+,K+-ATPase activity of the gills, a reduction in diffusive ion losses and a restoration of internal ionic level (Pelgrom et al., 1995; Li et al., 1998). In contrast to Hg and similar to Cd, there is no metallothionein induction in the gills or change in the gill copper uptake or binding (Grosell et al., 1996, 1997; Wood, 2001; Hollis et al., 2001).
More than 200 zinc based enzymes or other proteins have been identified from various organisms. In contrast to Cu (and iron), zinc does not form free radical ions and has antioxidant properties (Powell, 2000, cited by Bury et al., 2003). Zinc inhibits active Ca2+ uptake across the branchial epithelium and vice versa. One result may be hypocalcemia, since Ca2+
and Zn 2+ share a common uptake pathway via the chloride cells (Wood, 1992;
Hogstrand et al., 1998; Wood, 2001). In contrast to copper, zinc does not inhibit net Na+ and Cl- uptake in the gills, so that the plasma level of these ions is not affected.
Zn induces metallothionein in branchial tissue (Hogstrand et al., 1995). Chloride cell proliferation on the respiratory lamellae has been reported at high chronic zinc levels, explaining the long-term
increases in branchial Na+,K+-ATPase activity, and the repair of structural damage in the gills appears to proceed in the same way as described for copper (Watson and Beamish, 1980; Thomas et al., 1985; Spry and Wood, 1989;
Hogstrand et al., 1995; Wood, 2001).
Potassium:
Potassium is an essential mineral for animals due to its role in electrolyte and acid-base balance, which helps to maintain plasma viscosity and osmotic pressure (Emsley, 2003). Red blood cells contain the most potassium, followed by the muscles and brain tissue. The most important functions of K in the cells are:
regulating intracellular fluids, solubilising proteins, operating nerve impulses and contracting the muscles. Potassium, being a K+ ion, concentrates inside the cells, and 95% of the body’s K is located in the cells, unlike sodium and calcium, which are more abundant outside the cells. Cell membranes have channels for selective transport of Na+ and K+ ions against the concentration gradient (Hochachka and Somero, 2002). Information about the physiological effects of potassium on fish seems to be very sparse (Fielder et al., 2001).
Lithium:
The lethal concentration of dissolved Li for human ingestion is 5 g l-1 (Kszos and Stewart, 2003). Corbella and Vieta (2003) reviewed more than 300 studies on the long-term effects of Li on humans. The main effects of Li are neuropsychiatric indications, such as acute mania, depression, aggression, schizophrenia, and teratogenic (Ebstein’s anomaly) and metabolic effects. Lithium can stimulate glycogen synthesis in mammals through the activation of glycogen synthase and by mimicking insulin action, which causes weight gain. Lithium has teratogenic effects in amphibians (Boğa et al., 2000) and may affect embryonic development (Stachel et al., 1993). Lazou and Beis
12 (1993) found that Li affected the plasma membrane protein pattern in amphibian embryos. Bury et al. (2003) in their review of metal uptake in teleost fish supposed that Li passes via a putative epithelial sodium channel, which is probably also an important pathway in sodium-sensitive copper uptake (Grosell and Wood, 2002).
2.3. Effects of temperature, pH level, water hardness, and iron deposition on metal uptake and toxicity
The toxicity of metal ions depends on water quality, ion composition and the acid-base equivalents, as well as on the tolerance of the fish or other aquatic organisms, and digestive or respiratory metal uptake and excretion (Newman and Jagoe, 1994; Hamilton, 1995; Downs et al., 1998; Mason et al., 2000; Wood, 2001;
Mason, 2002; Duffy and Zhang, 2002).
The effects of abiotic factors, such as temperature, are of great importance as reflected in Table 1. In general, the higher the temperature of the water, the more toxic are metals such as Hg, Ni, Cr and K (Eisler, 1986; Fisher et al., 1991;
Hamilton, 1995; Mason et al., 2000;
Wood, 2001; Duffy and Zhang, 2002).
Campbell and Stokes (1985) described two contrasting responses of an organism to metal toxicity combined with a declining pH level:
1) a reduction in competition for binding sites with hydrogen ions (if the change in speciation is small and the metal binding is weak at the biological surface),
2) an increase in the metal availability (if there is a marked effect on speciation and strong binding of the metal at the biological surface).
In general, metals such Cd, Cu, Cr are more bioavailable and toxic for aquatic organisms in acidic water (Bradley and Sprague, 1985; Eisler, 1986; Lauren and McDonald, 1986; Playle et al., 1993;
Hamilton, 1995; Pelgrom et al., 1995;
Mason et al., 2000; Hollis et al., 2000;
Wood, 2001; Naddy et al., 2002; Bury et al., 2003).
The ion transport and permeability characteristics of fish gills are known to be extremely sensitive to water chemistry, particularly to hardness. Hardness is caused by main ions such as calcium (Ca) and magnesium (Mg), and for most toxicants, Ca2+ exerts a greater protective effect than Mg2+ (Wood, 2001). Calcium regulates the permeability and stability of membrane proteins in fish gills (Flik and Verbost, 1993). The higher Ca2+ (and Mg2+) concentrations in water lead to lower availability of metals. Similarly, a high level of dissolved organic carbon (DOC) reduces Hg, Cd, and Cu bioavailability (Playle et al., 1993;
Hamilton, 1995; Mason et al., 2000;
Niyogi and Wood, 2004). In addition, compounds such as chlorine ion and dissolved organic matter (DOM) have been found to limit Hg bioavailability (Newman and Jagoe, 1994). High water alkalinity reduces the toxicity of Cd, Cr and Cu (Carre et al., 1983; Linnik and Nabivanetch, 1986; Niyogi and Wood, 2004). High concentrations of sodium ions were found to reduce Cd and Li toxicity in water habitats (Kszos et al., 2003; Niyogi and Wood, 2004). Hardness is thus a surrogate for the protective aspects of water chemistry. It also has effects on the diffusive gradients for passive ion loss, substrate concentrations for active ion uptake, speciation on the toxicant or charged groups on the branchial surface (Wood, 2001).
It has been recognized that iron in the form of freshly precipitated ferric hydroxide Fe(OH)3 quickly scavenges many heavy metals, organics, and phosphorus (Jones Lee and Lee, 2005).
Martinez and McBride (2001) reported that Cu, Cd, Pb, and Zn were coprecipitated with ferric hydroxide, and that the binding of the metal to the ferric hydroxide depended on the type of metal, and for some metals, on the age of the precipitate.
13 2.4. Mechanism of metal toxicity in gills of freshwater organisms
The gills are the most vulnerable fish organs for metal toxicity. The extremely thin (0.5-10 µm) diffusion distance between water and blood and the effectiveness of the diffusive exchange, is maximised by the counter current flow of water versus blood at the exchange sites (Wood, 2001). The apical epithelial surfaces of gills are rich in the cells involved in ion exchange (pavement and chloride cells) and in cells that attract specific ions, such as water-borne metals, by their external mucous layer being charged negatively (mucous cells) or neutrally (rodlet cells) (Mallatt, 1985;
Matey, 1990; Reid and McDonald, 1991;
Wood, 2001; Evans et al., 2005). These facts explain why fish gills are the first sites of uptake for many water-borne toxicants such as metals.
The similarities and differences in metal toxicity mechanisms for freshwater fish gills are summarized in Table 1. The table is organized according to binding affinities (Log K) (Niyogi and Wood, 2004). Roughly, all other factors being the same, the Cd2+ is more toxic to gills than Cu2+, which in turn is more toxic than Zn2+
and so on. Consequently, the higher binding affinities (log K) bring greater
toxicity for a particular metal. Wood (2001) concluded that the toxic mechanisms at the gills have diversities for different contaminants at environmentally realistic concentrations, and therefore no general mechanism of toxicant action in the gills can be proposed. The main toxic site of metal actions in the gills is the Ca2+
membrane channel for Cd2+, Zn2+ and methylmercury. Copper acts through the Na+ channel, and Li+ is assumed to act similarly. Ionoregulatory toxicants, such as Cd2+, inhibited Ca2+ or Na+ and Cl- influx, as did Cu2+ and both forms of mercury.
These toxicants affect the enzymes Ca2+- or Na+, K+-ATPase activity. The latter enzyme appears to be a key toxic site for the Na+ antagonist. However, it was also inhibited by mercury, which acts (or is assumed to act for inorganic Hg) through the Ca2+ membrane channel. Thus it is possible to observe that the metabolic pathway does not depend on the toxic site of action. In other words, mercury can use both the Na+ and Ca2+ transport ion channels to penetrate the gill membranes (Playle, 1998).
Since abiotic factors have a high importance for metal availability (as described above), short notes are also given in Table 1 for each metal (for references see chapter 2.2.).
Table 1. Metal toxicity mechanisms for fish gills in freshwater.
Log K The point of toxic action in gills Metal uptake localisation Mechanism of action References Effects of abiotic factors on metal bioavailability Cd 8.6 Ca 2+ transport Chloridecells Inhibits Ca 2+ influx and Ca 2+-ATP activity, while bindingto calmodulin protein, no metallothionein induction. Wicklund-Glynn et al., 1994; Hollis, 1999, 2000; 2001; Wood, 2001; Niyogi and Wood, 2004 pH, alkalinity, DOC, Ca 2+, Mg 2+, Na +, Zn 2+
Cu 7.4-8.0 Na + transport, canbe Na + insensitive Chloride cells Ionoregulatory toxicant: inhibits Na + and Cl - influx, raises transcellular permeability, Na +,K +-ATPase inhibitor byhigh affinity to SH-group, stimulates mucous cell secretion. Lauren andMcDonald, 1985;Pelgromet al., 1995; Li et al. 1996; Wood, 2001, Grosell et al., 2004a,b pH, alkalinity, DOC, Ca 2+, Zn 2+**, Ni +**, Cu 2+**
Zn 5.3-5.5 Ca 2+ transport Chloridecells Inhibitor of carbonic anhydrase and Ca 2+-ATPase activity Hogstrandet al., 1995; Hogstrand and Wood, 1996; Wood, 2001 pH, Ca 2+, Mg 2+, Ni +**, Cu 2+**
Cr(III); Cr(VI) U/n U/n U/n More toxic at higher temperatures Eisler, 1986; Pawlisz et al., 1997 t oC*, pH, alkalinity, DOC, Ca 2+, Mg 2+, Ni+**, Cd2+** MeHg U/n Penetrates bydiffusion; Ca 2+
transport Chloride cells Ionoregulatory toxicant: Lipophilic, inhibits Na+ and Cl-influx, raises transcellular permeability; Na +,K +-ATPase inhibitor by high affinity to SH-group Lock et al., 1981; Heath, 1995; Playle, 1998; Wood, 2001 t oC*, DOC, DOM, Ca 2+, Mg 2+, Cl -, Se 2+
Hg U/n U/n, assumed Ca 2+
transport Distributed diffusely in gill epithelium Ionoregulatory toxicant: Permeability increase, inhibits Na + and Cl - influx, Na +,K +-ATPase inhibition, stimulates mucous cells proliferation and secretion Li U/n Reportedly by Na +
transport Assumed bychloride cells Assumed as an ionoregulatory toxicant. Water ions K +
and Na + play a protective role against Li Grosell and Wood, 2002; Kszos et al., 2003; V. Na +, K +
U/n – unknown; * the lower the factor the lower the metal bioavailability; **-additive toxicity effect
15 3. Material and methods
3.1 Field sampling
Three lakes in the Kenti river system – L.
Poppalijärvi, L. Kento and L. Upper Kuito were sampled in early August in 1997, 1998, 2000 and 2001. Some samples were from April 1997 and 2000 and some were taken from L. Kamennoe (in the year 1997 and 1998). L. Kamennoe and L. Upper Kuito are reference lakes not affected by mining waters. Samples for water quality were taken and analyzed according to Finnish standard methods (SFS Standards) in the water laboratory of the University of Joensuu. For all sediment samples only the first 4 cm of the surface were used. The sediments were stored frozen for 2 months before the analyses.
Multimesh gillnets were used to monitor the abundance and structure of the fish fauna (e.g. Kurkilahti and Rask, 1996).
Stratified random sampling was carried out in each lake with nine nets (3 inshore or 0- 3 m, 3 on the surface and 3 on the bottom offshore) for 12 h. The fish were identified, counted, measured and weighed according to species, and sampled for scales and opercula for growth analysis (Holopainen et al., 2003b; I). The samples for chemical and histological analyses were taken immediately after catching. Descriptions of the fish sampled from different lakes are given in (III), except for the fish from the year 2001. The results for these fish are unpublished and given in Table 2.
Table 2. Characteristics of fish sampled from the Kostomuksha area. Mean ± S.E., values with no common letter (superscript) are different when compared by ANOVA followed by the Post Hoc LSD-test (p<0.05).
Poppalijärvi Kento Upper Kuito
Summer 2001 August 2-3 August 8-9 August 1-2
Perch
n 15 15 15
Fork length (mm) 163±4a 175±5a 169±4a
Weight (g) 54±4a 97±25b 61±6a
Liver somatic index 0.64±0.05a 0.77±0.08a 0.97±0.06b
Roach
n 15 15 15
Fork length (mm) 173±3a 166±3b 143±5c
Weight (g) 79±5a 69±5a 41±4b
Liver somatic index 0.63±0.06a 0.94±0.16a 0.8±0.06a
3.2. Laboratory experiments
The effects of a Li&K mixture (Experiment I) and Li only (Experiment II) were studied in juvenile rainbow trout (Oncorhynchus mykiss). This is a fish species with a high metabolic level suitable for adequate response, even in a short-term laboratory experiment. In both experiments rainbow trout were obtained from a local
fish farm, brought to the laboratory and acclimated to the local tap water with comparable pH and oxygenated levels (IV, V). The fish were fed with granulated industrial dry food to an amount of about 1% of fish mass. During the experiments the fish were not fed. The total fasting period in both experiments was three weeks.
16 The differences in the experiments were water hardness (soft in the Li&K mix and medium hard in the Li exposure), temperature (12 and 13oC in the Li&K mix; between 14 and 15oC in the Li exposure), standing water (Li&K) and flowing water (Li), fish weight, and Li
concentration (Table 3). Nevertheless, concentrations of Li were comparable between the exposures and similar to the condition of L. Poppalijärvi regarding the Na+/Li+ log ratio, and fish sizes were also maximally comparable (Table 3.).
Table 3. Water and fish characteristics during the laboratory exposure. Rainbow trout were exposed to the Li&K mixture in standing water (Feb. 2002, Joensuu) and in flowing water to the Li only (a step-up exposure, July-August, 2004, Hamilton). Mean ±S.E.
ToC Ca
mg l-1 Mg mg l-1
Liµg l-1 Na mg l-1
K mg l-
1
Log Na/Li
Log K/Li
Weight , g
Fork length mm Exp.I 12-
13 16 1.9-
2.6 27±4 2.3-
2.9 115-
126 0.35 -0.2 19±1 11.9±0.2
Exp.II 14-
15 40 4.9 66±27;
528±165
13.6 2 0.07;
0.2
1.2;
1.2 13±2 10.5±0.5
3.3. Analyses
3.3.1. Heavy metals
After taking fork length (AC) measurements and weights of all the fish caught from each lake, 15 fish were randomly selected. Samples of skeletal muscle (2-3 g) were dissected from the left side (between the dorsal fin and lateral line) of each fish. The whole gill arches and whole livers were dissected (with hands dressed in powderless surgical gloves). Each tissue sample was collected into a separate uncoloured polyethylene vial. Before sampling, each vial was cleaned in 10% HNO3 for 5 hours and washed thoroughly with de-ionized water, rinsed 3 times, dried, labeled, preweighed and packed in two clean paper boxes with lids. Samples for metals, histology, enzymes, and lipid analyses were taken from each fish simultaneously and placed in separate vials. The samples were kept on ice (1997, 1998) and in liquid nitrogen (2000, 2001) until arrival at the laboratory.
Samples were kept in the freezer for up to two days (-20oC, samples from 1997 and 1998) and for up to two months (-20oC inorganic samples from 2000 and 2001).
Then the samples were dried at 105°C for
about 12 hours. The dry samples were kept in the same vials at +4°C until they were analyzed less than 6 months later (samples from the years 1997-1998) or analyzed immediately (all other samples). The dried samples were digested in a microwave digestion unit (Milestone 1200 mega) in a mixture of HNO3 and H2O2 in proportions of 4:1 for fish, and 5:1 for sediment.
Cadmium, chromium and nickel in water, sediment and fish samples were analyzed by graphite furnace AAS (Hitachi Z-9000).
Copper and zinc were measured by flame AAS and mercury concentrations by a gold-film mercury analyzer (Jerome Inst.
Corp. Model S-11). Water potassium and Li concentrations were analyzed by atomic absorption spectrometry (Perkin Elmer AAS). Minimum detection limits (ng ml-1) according to the technical notes were:
Cd=0.0014; Cr=0.0038; Ni=0.072; Zn=0.8;
Cu=1; Hg=0.009. The reagents blank value and standard solutions with known concentration were used for the quality control.
3.3.2. Histology and histochemical analyses
A transmission electron microscope was used to study the structure of the liver (II)
17 and the gills (III, IV) in Epon section.
Light microscopy was used to count iron deposits in the gills after histochemical staining in paraffin section (III).
3.3.3. Enzyme analyses
The field and Experiment I: total ATPase and Na+,K+-ATPase activities in the gill were measured using a kinetic method described by Johnson et al. (1977).
Proteins were measured by standard methods with the Folin phenol reagent (Lowry et al.1951). In the gills of Li- exposed fish (V), total ATPase and Na+,K+-ATPase activities were measured using a kinetic microassay run in 96-well microplates, as outlined by McCormick (1993). The latter method is preferred because it requires less sample tissue and chemicals. However, the first method measured the total ATPase activity (III, IV), which is not the case in the second method (V).
The citrate synthase (CS) activity was measured using a kinetic microassay run in 96-well microplates as described by Leonard and McCormick (1999). Protein concentrations were determined by the Bradford method (Biorad) (V).
3.3.4. Lipid analysis
All lipids (about 100 mg) were extracted from the gill tissue (separated from the cartilaginous arches, left part of the gills) and homogenized in a chloroform- methanol (2:1, vol/vol) solution (Blind and Dyer, 1959). Then the lipids were extracted at room temperature during 40 minutes in the dark and collected as described in articles III, IV and V.
However, after the extraction, the lipids were redissolved and analyzed differently (III-V).
A thin layer chromatography, flame ionization detection system (Iatroscan, Iatron Laboratories Inc., Japan) was used at the University of Joensuu (Finland) for gill lipid determination. All the extracts
(samples, standard, blank) were scanned in duplicate and then calculated in relation to the internal standard (cholesterol acetate) (IV and unpublished data in the Summary).
Since Iatroscan was not available at McMaster University (Canada), we used enzymatic colorimetric methods for the quantitative determination of total free fatty acids and total cholesterol with commercial diagnostic kits (NFFA C and Cho E, Wako Chem.) (V). The latter method is less time-consuming, but it limited the diversity of lipids to those for which the kit was designed.
3.3.5. Osmolality and ionic concentrations of plasma
Blood samples were obtained only in Experiment II. Plasma measurement of osmolality, total Na+, K+ and Cl- concentrations were made in order to assess the effect of Li exposure on juvenile rainbow trout (V).
3.3.6. Apolipoprotein AI analysis
A semi-quantitative Enzyme Linked Immunosorbent Assay (ELISA) was used to assess ApoAI in blood plasma, since purified trout apolipoprotein AI (apoAI) was not commercially available for a fully quantitative assay. The apoAI was analyzed in plasma samples from the second step of exposure at the higher Li concentration. A detailed description of the method is given in article V.
3.3.7. Statistics
All heavy metal concentrations in tissues (µg g-1) are expressed on the basis of dry weight. The heavy metal data were log10
transformed prior to statistical analysis. A non-parametric Kruskal-Wallis test used to compare three independent groups of sampled data (Zar, 1999). For statistical similarities and differences, and to find out whether these are due to the effect of metals or are changes depending on time or the lake in question, fish morphometric
18 parameters, concentrations of plasma ions, osmolality, lipids and enzymes values were compared by one-way analysis of variance (ANOVA), followed by the Least Significant Difference test (LSD). The Kolmogorov-Smirnov test was used in both the exposed and the control group for determining statistical significance as against Li added directly to the media for Na+,K+-ATPase analyses. The Student’s unpaired two-tailed t-test was performed with control fish and gill lipids sampled at the same time from L. Poppalijärvi and L.
U. Kuito. The Student’s t-test was used for 405 nm absorbance at each dilution in the ELISA assay to compare the mean area and describe the effect of Li along the plasma dilution curve. SPSS 10.0 and Excel 2000 computer programs were used for statistical analyses.
4. Results
4.1. Fish communities
Seven species of fish were caught in L.
Poppalijärvi and L. Kento. In the reference lake the number of species was eight; dace (Leuciscus leuciscus) was missing from the impacted lakes (I). Roach and perch were most numerous, and roach were predominant in the catch from all lakes.
The total biomass per unit effort was twice as high in the impacted lakes (Holopainen et al., 2003b; I).
4.2. Metals in the water and sediment Table 4 summarizes the water quality data from the field studies in the years 1997-
2001. Copper shows a higher concentration in L. Poppalijärvi in the spring than in the summer. Chromium was higher in the uppermost L. Ahvenjärvi than in L.
Poppalijärvi or in the other lakes. High concentrations of Cr (91 µg l-1) were found in the wastewater data of Virtanen and Markkanen (2000). Cadmium primarily affected L. Poppalijärvi.
Aluminium, Zn and Fe were low in the L. Poppalijärvi water when compared to the reference lake, L. Upper Kuito (Table 4). Iron content fluctuated a lot in summer and early spring, and was, for example, about ten times higher in early April 2000.
The alkali metals Li and K were found at much higher concentrations in L.
Poppalijärvi compared to the reference lake (Table 4).
Sediment analysis demonstrated high concentrations of Hg, Cd, Zn, Cu and Fe in L. Poppalijärvi in 2000 (Table 5). The reference lake had a higher value for Cr deposition than the study lakes. The sediments of L. Kento appear to have been contaminated more heavily in the past since the top layer had a low metal concentration (Table 5), except for Fe. Iron showed a higher concentration in the surface layer (0-2cm) than in the deeper layer (2-4 cm). While the average concentrations of Cu and Cr were low, the Hg, Cd, Fe and Zn showed elevated concentrations in the studied lakes.
19
Table 4. Lake description and surface water quality in the Kenti river system in 1997-2001.
Year 1997 – data from Morozov (1998) and Virtanen and Markkanen (2000). A – results from L. Ahvenjärvi. No asterisk – data from August 2000 and 2001, * data from April 2000, **data from Lozovik et al., 2001 (L. Middle Kuito), ***data from August 2002.
Lakes from the Kenti River System Reference lakes
Poppalijärvi Kento Kamennoe Upper Kuito
Surface area (km2) 1.7 27.1 95.5 197.6
Maximum depth, m 11 24 29 44
Year 1997 2000 2001 1997 2000 2001 1997 2000 2001
pH 7.6-
8.2 8.3 8.4 6.6 7.7 7.7 6.4-7.0 6.8 6.7
Oxygen, surf./ bot., % of saturation ***
107/95 94/97 97/74 96/85 115/95 90/94
Conduct. mS m-1 - 39.1 39.4 - 14.9 14.0 2.6 2.5 2.4
Color, Pt mg-1 50 35 (50*)
- 60 40 - 25 40
(70*) -
TOC mg l-1 6.2 7.7* - 7.2 - - 6.1 11.6* -
IC mg l-1 - 17.8* - - - - - 2.3* -
Tot. Phosph. µg l-1 8 6 9 7 8 8 7 14 11
Tot. Nitrogen, µg l-1 2200 3888 4038 270 797 697 220-350** 263 581
SO42- mg l-1 56 82.1 - 23 20.6 - 2 2.2 3.6**
Cl- mg l-1 3.7 3.7 - 1.5 1.5 - 0.8 0.8 -
Na mg l-1 6.0 6.2 (8*) 6.4 2.0 2.8 2.6 1.2 1.2
(2.1*) 0.8
K mg l-1 60 60
(62*)
60 13 20 20 0.4 0.5
(1.0*) 0.4
Li µg l-1 17-38 20 - 7 7* - 0.2 0.2 -
Ca mg l-1 35 21
(18*)
23 6.2 7.9 8.4 1.6 1.9
(1.6*) 1.9
Mg mg l-1 8.0** 8.3 7.7 3.0** 3.1 3.2 - 0.8 0.6
Al µg l-1 5.9A 11 - - 17 - - 57 35-
60**
Zn µg l-1 0.63A 6 (8*) - 0.45 9 - 5 16 (0*) -
Cu µg l-1 0.49A 4.31* - 0.33 - - 1 0.93* -
Cr µg l-1 8.3A 0.38* - 0.20 - - <1.0 0.37* 0.55-
0.58**
Cd µg l-1 <0.03 0.20
(0.19*)
- <0.03 0.1 - <0.03 0.1 (0.07*)
-
Fe µg l-1 ≈600 38-45
(434*) 58 ≈200 98-
104 123 300 202-
266 (269*)
240
